---
title: "Redaction of sensitive data"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Redaction of sensitive data}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

## Introduction

GEDCOM files are often shared between people, but a file can contain detailed information for individuals, which is an issue, especially if they are still alive.

The `tidyged.utils` package contains functionality to detect these individuals and either remove them or remove their details.

## Basic usage

The function to remove living individuals is `remove_living()`. To illustrate, we create a tidyged object containing a number of different individuals, alive, dead, and some ambiguous:

* A person born in 1996
* A person with a death event
* A person born in 1796
* A person who was a driver in 1930 when they were 50 years old
* A person who got married in 1856 when they were 20 years old

```{r}
library(tidyged)
library(tidyged.utils)

people <- gedcom(subm("Me")) |> 
  add_indi(qn = "Living person") |> 
  add_indi_fact("birth", date = date_calendar(1996)) |>
  add_indi(qn = "Confirmed dead person") |> 
  add_indi_fact("death") |> 
  add_indi(qn = "Reeeaally old person") |> 
  add_indi_fact("birth", date = date_calendar(1796)) |> 
  add_indi(qn = "Implicit dead person 1") |> 
  add_indi_fact("occupation", descriptor = "Driver", date = date_calendar(1930), age = "50y") |> 
  add_indi(qn = "Implicit dead person 2")

idp2_xref <- find_indi_name(people, "Implicit dead person 2")

people <- people |>
  add_famg(husband = idp2_xref) |> 
  add_famg_event("relationship", date = date_calendar(1856), husband_age = "20y")

describe_records(people, people$record)
```

The default behaviour of the function is to remove data for living individuals, but also those that are ambiguous. Confirmed dead person has a death event, and Reeeaally old person was born in 1796. The function assumes a maximum age of 120.

```{r}
remove_living(people) |> 
  describe_records(people$record)
```

For illustration purposes, we can increase the maximum age threshold, which will make the function treat the old person as still living:

```{r}
remove_living(people, max_age = 300) |> 
  describe_records(people$record)
```
## Guessing individuals ages

The `guess` parameter will cause the function to invoke additional functionality to try to guess the age of individuals where a date of birth is not given:

```{r}
remove_living(people, guess = TRUE) |> 
  describe_records(people$record)
```

This causes the function to determine that all individuals apart from the first is dead. This looks in both individual facts and family group events.

## Redaction options

The remaining parameters determine the action to take when living individuals are found. By default, the records are preserved, but all detail is removed, leaving only a change date and an explanatory note:

```{r}
remove_living(people, guess = TRUE) |>
  dplyr::filter(record  == "@I1@")
```

The user has the option of changing the text of this note using the `explan_note` parameter. Alternatively it can be set to an empty string to remove it completely:

```{r}
remove_living(people, guess = TRUE, explan_note = "") |>
  dplyr::filter(record  == "@I1@")
```

Alternatively, the record can be removed completely:

```{r}
living_removed <- remove_living(people, guess = TRUE, remove_record = TRUE)
```

Since supporting records could also hold sensitive information, the `remove_supp_records` parameter allows you to also remove these (which it does by default).